Research On Fault Diagnosis Expert System With Knowledge Database For Diesel Engine

Diesel engine is the primary power plant in the warship, which performance directly effects navy's battle effectiveness. Diesel engine fault is one of the crucial reasons on the sea accident. So fault diagnosis technology for marine diesel engine is regarded by navy and shipping community all the time. Because of the complexity of diesel engine structure, the traditional diagnosis method for the rotary machine as research object, couldn't deal with these problems effectively.This paper, with 16VPA6 marine diesel engine as the research object, adopts the fault diagnosis technology with popular expert system. Man-machine interface is designed with interface developing tool Borland C++ Builder and database software SQL Server. The practical fault diagnosis expert system for 16VPA6 marine diesel engine is developed according to the design ideals and steps of expert system.Following aspects are the most important tasks in this paper. Firstly, the entire function module and main interface of expert system are designed after analyzing the user's requirement. Secondly, the knowledge representation mode, acquisition approach and the reasoning method are confirmed through the research on the knowledge database. The knowledge database of expert system including the fact database, rule database, advice database, initial database, background knowledge and database for saving parameters from sensors are established according to the data collected. On this basis, the knowledge database functions including management function and dealing with system error are improved, and corresponding data table is built. Based on the fault tree, self-learning module is constituted. Thirdly, some assistant modules of expert system including user login, register and management, bilingual interface switch are established. Fourthly, according to the gray system theory, some parameters for long-time or short-time are forecasted, meanwhile parameters of the diesel engine system including fuel systems, lubrication oil systems, cooling systems, boosting systems and start-up systems will be on-line monitored and alarmed. Fifty, reasoning mechanism and interpretation mechanism are determined, and the diesel engine system can then be monitored and diagnosed. The main diagnosis methods adopted are respectively based on the trend analysis of thermodynamic parameters such as temperature, pressure and speed of the main engine, the regular analysis of the indicator diagram, in which the processed data are compared with the normal curve with LABVIEW as pre-processor, the abstraction of fault information in vibration signals through the wave diagram, frequency spectrum, power spectrum and quefrency spectrum.The trend of the parameters for PA6 diesel engine systems can be monitored and forecasted by the expert system designed in this paper, also some faults can be diagnosed and the repair proposals can be come up with. The function of expert system will be strengthened in the subsequent use and learning.